How can I change the python version on a Watson Studio Jupyter notebook?

Question

I'm working on Watson Studio Jupyter notebooks, and currently the default is python 3.9 or 3.8. My notebook requires tensorflow 1.15 because I'm using R-CNN from this project https://github.com/matterport/Mask_RCNN/releases. However, tensorflow 1.15 is incompatible with python>=3.8. Thanks!!

Answer 1

IBM Watson Studio runtimes do not currently (as of this writing) support any Python versions other than 3.8 and 3.9, as illustrated in IBM's official docs on Compute resource options for the notebook editor in projects.

More specifically, support for the next-oldest Python version beyond 3.8 (which was consequentially Python 3.7) was deprecated on 25 Feb 2022 and subsequently unavailable for new IBM Watson Studio projects/deployments - see What's new page:

Support for Python 3.9 and deprecation of Python 3.7 (Watson Studio and Watson Machine Learning)

You can now use IBM Runtime 22.1, which includes the latest data science frameworks on Python 3.9, to run Watson Studio Jupyter notebooks, train models, and run Watson Machine Learning deployments. Python 3.7 is now deprecated and will be removed on Apr 14, 2022. [...]

Similarly, Python 3.6 was removed from Watson Studio previously due to a security vulnerability. From the aforementioned What's new page:

Reminder: Removal of Python 3.6 environment (Watson Studio and Watson Machine Learning)

Python 3.6 is being removed from Watson Studio and Watson Machine Learning due to a security vulnerability. [...]

---

As such, your options are to either (a) change your architecture within IBM Cloud to remove your dependency on Watson Studio, (b) migrate to another cloud provider which has support for the versions of Python you require, or (c) (and likely the "correct" way) update your code to not depend on such old dependencies.